As the most common and deadly malignant primary brain tumor, glioblastoma (GBM) is almost universally fatal, with a 5-year survival rate of less than 5%. This statistic has not improved significantly in decades, and there is still no treatment option to preferentially target the glioma stem cells (GSCs) or diffuse infiltrative cells that lea to tumor recurrence after surgery, chemo or radiotherapy. Irreversible electroporation has demonstrated promise as a tool for the non-thermal ablation of tumor tissue, which may prove useful for the treatment of GBM. The non-thermal nature of IRE spares major blood vessels and leaves the extracellular matrix intact while destroying the cells within a specified volume. A challenge with this technique is the non-specific targeting of normal and malignant tissue alike. This lack of specificity is typically acceptable for treatment applications in the liver, lung, or ther major organs. However, translation of this approach for the treatment of GBM will require the development of protocols that specifically target malignant cells. Despite the fact that cellular morphologic alterations (e.g. nuclear size and shape) were among the earliest clinical characterizations of tumor cells, and remain a primary prognostic tool used by pathologists today, these characteristics have never been used to target anti-tumor therapy. We hypothesize that GBM cell ablation resulting from a high frequency targeted electric field (TEF) version of IRE depends on cell type, local microenvironment, and molecular pre-conditioning that may be used to enhance targeting efficacy. Our preliminary experiments and mathematical modeling suggest that by delivering a burst of pulses, in the range of 0.5 to 10 ?s, we can achieve full tumor ablation within a central tissue zone, while extending damage to only the malignant cells in the diffuse infiltrative niche due to the increased nuclear-to-cytoplasm ratio (NCR) of malignant cells. We will proceed in this research by pursuing two synergistic paths: 1) establish the role of cell type, microenvironment and molecular pre-conditioning via GBM-specific EphA2 activation in regulating the response of normal astrocytes, bulk GBM cells and patient-derived GSCs to TEFs, using miniaturized brain mimetic niche models with tunable microenvironment, and 2) Demonstrate preferential TEF targeting penetration into the diffuse infiltrative niche of GBM, using a novel GSC-based microfluidic model of this heterogeneous infiltrative microenvironment. The results of this project will lay the groundwork for a subsequent R01 to develop targeted therapies for the GBM infiltrative niche in animal models, which can then be translated into effective treatments for human GBM, as well as other complex and aggressive malignancies.